Identification of an intronic splicing enhancer essential for the inclusion of FGFR2 exon IIIc

The ligand specificity of fibroblast growth factor receptor 2 (FGFR2) is determined by the alternative splicing of exons 8 (IIIb) or 9 (IIIc). Exon IIIb is included in epithelial cells, whereas exon IIIc is included in mesenchymal cells. Although a number of cis elements and trans factors have been identified that play a role in exon IIIb inclusion in epithelium, little is known about the activation of exon IIIc in mesenchyme. We report here the identification of a splicing enhancer required for IIIc inclusion. This 24-nucleotide (nt) downstream intronic splicing enhancer (DISE) is located within intron 9 immediately downstream of exon IIIc. DISE was able to activate the inclusion of heterologous exons rat FGFR2 IIIb and human beta-globin exon 2 in cell lines from different tissues and species and also in HeLa cell nuclear extracts in vitro. DISE was capable of replacing the intronic activator sequence 1 (IAS1), a known IIIb splicing enhancer and vice versa. This fact, together with the requirement for DISE to be close to the 5'-splice site and the ability of DISE to promote binding of U1 snRNP, suggested that IAS1 and DISE belong to the same class of cis-acting elements.     

Identification of a conserved 125 base-pair Hb9 enhancer that specifies gene expression to spinal motor neurons

The homeobox gene Hb9 is expressed selectively by motor neurons (MNs) in the developing CNS. Previous studies have identified a 9-kb 5' fragment of the mouse Hb9 gene that is sufficient to direct gene expression to spinal MNs in vivo. Here, we sought to identify more discrete MN-specifying elements, using homology searches between genomic sequences of evolutionarily distant species. Based on homology screening of the mouse and human Hb9 promoters, we identified a 3.6-kb Hb9 enhancer that proved sufficient to drive MN-specific lacZ expression. We then compared mouse, human, and pufferfish (Fugu rubripes) genomic sequences, and identified a conserved 438-bp sequence, consisting of noncontiguous 313-bp and 125-bp fragments, residing within the 3.6-kb Hb9 enhancer. The zebrafish (Danio rerio) Hb9 genomic region was then found to have two identical copies of the 125-bp sequence, but no counterpart for the 313-bp sequence. Transgenic analysis showed that the 125-bp alone was both necessary and sufficient to direct spinal MN-specific lacZ expression, whereas the 313-bp sequence had no such enhancer activity. Moreover, the 125-bp Hb9 enhancer was found to harbor two Hox/Pbx consensus-binding sequences, mutations of which completely disrupted thoracolumbar Hb9 expression. These data suggest that Hox/Pbx plays a critical role in the segmental specification of spinal MNs. Together, these results indicate that the molecular pathways regulating Hb9 expression are evolutionarily conserved, and that MN-specific gene expression may be directed and achieved using a small 125-bp 5' enhancer.